‘Ensuring that services are always available and secure’
Laurens van der Spek, COO, and Jeroen van der Linde, Senior Technical Support Engineer, discuss the importance of a secure, robust and scalable IT landscape. It’s what forms the foundation allowing our algorithms and service to perform optimally. Autonomous growing requires the highest level of quality and should always keep running. They also reveal the new cloud technologies being applied, and how Blue Radix ensures the privacy and security of customer data.
What are you responsible for in Blue Radix?
Jeroen (on the right): “As a Senior Technical Support Engineer I’m responsible for everything going on in Azure. That’s the Microsoft platform on which our systems run. I monitor and manage our virtual systems in the cloud. The software we create ourselves also runs on Azure. The growth in customers, the volume of data and the importance of the systems for our customers all mean we need to have our systems scaled accordingly to ensure that they all keep running properly. This is also my responsibility.”
Laurens (left in photo): “As Chief Operations Officer I bear final responsibility for all the operational services for our customers, both functionally and technically. That includes our Crop Controller service, including the guidance of our customers and monitoring by the Autonomous Greenhouse Manager. I’m also responsible for all the IT developments and technical management. That could include the architecture of our IT landscape, security and privacy, along with managing the development teams who evolve new software, among other things. We’re constantly working together to discover new techniques and developments to help us further, while also guaranteeing stability.”
Blue Radix uses fairly new cloud technologies; why was this chosen and what are the challenges?
“The main reason behind working with the latest cloud technologies is because they’re better and easier to scale. You used to need extra capacity on a server, which took a lot of time and effort, while scaling up or down in the cloud is far easier these days. If set up properly it can even happen automatically,” notes Jeroen. “The management and security are part of the Azure platform, so you can develop services faster,” adds Laurens. “The downside is that it still takes lots of time and expertise to deal with this properly. Many new techniques and tooling are added every day, but software is also continuously being withdrawn. You have to keep up and choose the right new tooling. Not everything works for our objectives. We regularly carry out this assessment with the development teams. The biggest challenge of developing in the cloud is that while it can be done quickly, it can also head in the wrong direction. You have to consider the design of your IT landscape and architecture carefully in advance. That’s one thing we’ve learned in our years of developing and managing data-driven services at AgroEnergy – thinking ahead about the possible future, so you can respond to new developments. It’s very important that you keep following the right target architecture and develop within it. That lets you lay an essential foundation that will last for years.”
Jeroen: “The downside of scaling up systems on Azure easily and quickly is the cost. It’s important to keep striking the right balance between what you need and the costs this involves. That applies to upscaling, but you can also scale down; you can actually compare it to the data bundle on your cellphone.”
Is specific technical management still needed with the arrival of new AI and cloud technologies?
“Yes it really is needed, because we provide hourly services to growers 24/7. Crop Controller controls the installations in their greenhouses continuously. So it has to keep running all the time,” Laurens points out.
“Previously Technical Management used to manage three or four servers, and you knew everything there was to know about them. Nowadays it’s a whole chain of data sources, algorithms, models and applications, all of which have to be properly connected and linked. You have to know the ins and outs of this. Intervening and coordinating correctly is even more important now. There are many more relationships and effects involved in certain changes and adjustments.” Jeroen: “Fortunately we have in-house expertise, and we monitor the technical management continuously. We prefer to prevent problems proactively rather than solving them afterwards. A major advantage is that the chance of failure, or downtime, is virtually zero.”
How do you ensure the security and privacy of sensitive customer and company data?
“We work with a lot of our customers’ business-sensitive data, so we need to think carefully about just how to configure our security and safety,” explains Laurens. “We’ve set things up such that we can work very securely indeed with our customers’ business-sensitive data.” Jeroen: “One example is that only computers authorized by us have access to our systems in Azure. Additionally every employee uses two-way verification, meaning there’s a confirmation of your password via another channel, for example by cellphone. This two-way approval means you also need to change a password less often.”
Laurens adds: “Every month we receive information from Microsoft on whether, and where, we can improve our security. When working with partners we also ensure we have a Non-Disclosure Agreement (NDA) in place. We don’t share any customer data unless the customer has provided explicit permission. Regular updates, security audits and security training are standard procedures for us. We’re also affiliated with the Glass 4.0 data protocol. And of course the European Union has strict laws on privacy. We make every effort to comply with all these regulations, and have prepared a privacy and data statement to this end; it can be viewed on our website.”
How do you guarantee that all customer processes will continue being technically viable 24/7?
Laurens: “We have a monitoring team of colleagues from a range of fields who monitor the systems and the technical chain of applications seven days a week. In combination with the techniques we use, this ensures there’s only a minimal chance that something will fail.” But suppose the customer’s internet connection goes down? “Then the system will simply keep running locally. With autonomous growing we predict ahead, several days in advance. So if the internet goes down control continues as usual. Of course we’ll be in touch with the customer to resolve the problem. Then afterwards we retrieve the data from the climate computers, so there are no gaps in the data, and the historical data is guaranteed. Our job is to ensure that the algorithms keep running. They are the house, as it were, and we ensure that this house has the right foundations.”